skills.rest
Gamezar
GamezarCOMMUNITY
Explore all skills by this author →
GitHub

sadd-judge

Launch an isolated sub-agent judge to score conversation work with evidence-based rubrics.

Updated Mar 10, 2026
Search Tags:#quality assurance#code review#feedback#llm-as-judge#agent#evaluation
One-click install
npx skills add https://github.com/Gamezar/opencode-cek --skill sadd-judge-gamezar
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sadd-judge
Source: https://github.com/Gamezar/opencode-cek/tree/main/plugins/sadd/skills/sadd-judge
Command: npx skills add https://github.com/Gamezar/opencode-cek --skill sadd-judge-gamezar

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill automates the evaluation of work produced within a conversation, providing objective, evidence-based feedback to ensure quality and adherence to requirements.

Core Features & Use Cases

  • LLM-as-Judge Pattern: Leverages a specialized sub-agent for unbiased assessment.
  • Context Isolation: Ensures the judge operates with a clean slate, preventing bias.
  • Structured Evaluation: Provides scores with detailed justifications and evidence.
  • Use Case: After an agent generates code, use this skill to have a judge evaluate its quality, instruction following, and completeness against the original request.

Quick Start

Use the sadd-judge skill to evaluate the code generated in the previous turn.

Frequently Asked Questions about sadd-judge

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How does the LLM-as-Judge pattern work for code review?▼

Yes, an AI judge can evaluate code quality by operating with isolated context to prevent bias. It assesses instruction following and completeness against original requirements, providing detailed justifications and evidence for its scores.

How do I automate quality assurance evaluation for work produced in a conversation?▼

You automate quality assurance by launching a specialized sub-agent judge to evaluate the work produced in the current conversation. It operates with a clean slate and applies self-verification to deliver structured, actionable feedback.

Why use context isolation when evaluating agent-generated work?▼

Context isolation ensures the judge operates with a clean slate, preventing bias from the previous conversation turns. This allows the AI judge to provide an unbiased, objective assessment of the generated work against the rubric.

Can I use an AI judge to check if generated code meets original requirements?▼

Yes, you can use this evaluation skill to have a judge evaluate code quality, instruction following, and completeness against the original request. It implements self-verification to ensure the assessment is accurate and actionable.

What are the limitations of using an automated sub-agent for code evaluation?▼

The main limitation is that the sub-agent judge operates with isolated context, meaning it only evaluates the work produced in the current conversation. It requires clear multi-dimensional rubrics to deliver structured, evidence-based scoring effectively.

Related Skills

Explore More Software Engineering Skills→
m2ai-portfoliom2ai-portfolio
4

recognizable-quality-test-generator

Generate a structured quality rubric with pass/fail criteria and calibration steps.

Official
Advanced
irahardiantoirahardianto

debugging-protocol

Identify root causes of software issues through a structured debugging workflow.

Community
Advanced
viniruggeriviniruggeri

scholar-evaluation

Assess scholarly work across eight ScholarEval dimensions with scoring and feedback.

Community
Advanced
luicabref97luicabref97

sadd:judge

Generate YAML rubrics and score conversational artifacts with cited evidence.

Community
Advanced